The h Bar

Swords are scary things. Even in incompetent hands.

@tpg2114 It wouldn't have draw a batted eye in Santa Barbara either.

The Tai Chi guy held weapons practice on the Chancellor's Lawn every week.

@dmckee We have DragonCon and MomoCon take over campus during various breaks when the conferences are in town. People running around in costume with weapons is pretty normal

12:53 AM

@tpg2114 There's an East-Asian institute on campus here, with quite a collection of...*artifacts*. You wouldn't even have to bring your own sword, I guess.

Swords, glaives, spears...

But you had to have approval before you could come to that workout.

Though you'd have to have the desire to slice someone up during its opening hours...

user54412

@bolbteppa physics.stackexchange.com/questions/401/… might be a good place to start

bolbteppa

Haha thanks but I don't think such a program will show me how to determine the projections of the ellipsoidal coordinates onto Cartesian axes, I don't think anybody has ever done it :(

user54412

ah yeah - I can't really help there

12:59 AM

Science is hard

3

@tpg2114 That's why I'm doing my best to leave it

@KyleKanos What do you have in mind?

Quantitative Finance

All the alternatives that interest me are also hard.

I have an in over at Wells Fargo in Charlotte

1:03 AM

I spent all weekend writing python scripts using VTK to do some analysis, it's taking forever to run and it's very sensitive to the choice of parameters

Yeah, that's not going to be any easier, but they say the paycheck is sweet.

I have to learn stochastic calculus by December

And C++

@KyleKanos Good luck... I tried to take that class, I didn't make it 2 lectures before it was so far beyond me I dropped it

Well, hit the books about the math, and you poor bastard about c++.

The professor said "This is easily the most difficult math class that is offered"

1:04 AM

Do they at least give you a specified subset to understand?

@tpg2114 I'm registered for an "intro" course on that (400 level), plus I've been studying on my on

@tpg2114 FSM! Then I don't want to know about it. Coordinate-free, differential geometry did for me.

@dmckee No, but I know that's what they use

@KyleKanos I jumped into the graduate level class and I didn't have any classes in real analysis prior to it... bad idea

That's a problem because it is a big language and there are a lot of different ways to use/approach it.

1:07 AM

The language is small... the disgusting number of libraries make it overwhelming. I spent like 2 months trying to figure out what linear algebra library to use

And I ended up just picking one and running with it pretty much arbitrarily

@tpg2114 With meta-programming and all the 2011 extensions I don't think it is fair to characterize it as small. I admit it is not Ada...

I'll just have to teach them that Fortran is superior, that's all

C is a nice small language. With a nice small book. Except that there isn't a completely up to date version of K&R anymore (and now R is gone to the big free access memory in the sky).

Augh. If an author has a seemingly random $a_z$ in an equation, what would you guys assume it was?

@AmagicalFishy What subject?

1:10 AM

E&M, it's an equation for the magnetic field

B_0 sin(wt) a_z

@AmagicalFishy The real coefficient of the zth term. Unless it's QM in which case it is a complex coefficient.

user54412

wonder if that's supposed to be \hat{e}_z

I think that Chris may have the right answer to this instance, though.

That's what I was thinking, but he's got an a_z^2 in the (not-very-explained) step in between the presentation and the answer

(I'm real bad at E&M and trying to go through this book, 2008+ solved E&M problems, to ... not suck)

(The downside of having so many problems so packed together is that some of the answers are difficult for me to follow)

I'd need more context. Look closely at the question and any accompanying diagrams, of course.

1:16 AM

I'm going to go through it again and see if I can figure it out; if anyone's curious, I'll report back

Side thought on the "carry swords across campus" bit. I was recalling a time a couple of decades ago. I doubt you'd be allowed to get away with that most place these days.

I mean, all the building at my current place and the national lab I go to say that "Dangerous Weapons" are forbidden.

And I keep wondering about the not so dangerous kind.

@dmckee Ever break a CD in half? Sharper than most knives. Anything can be a weapon, and non-weapons can be dangerous

Not so dangerous weapons... mace (the spray, not a spiked ball on a stick)? Maybe non-lethal and self-defense weapons?

That's all I can come up with though, and I bet those would still be forbidden

@tpg2114 The security professional know that. And any amateurs who bother to learn anything and think about it, but neither group set policy or writes the silly signs.

I sometimes carry an umbrella that could be put into service as a baton. Or a cane--I plead a trick knee that gives out unexpectedly from time to time, which was true for a while but hasn't happened in years..

And my fitness routine is martial arts, because I need there to be other people expected me or I skip too much.

But against a sword all that is pretty weak, and against a gun? You'll need luck to even have a fighting chance.

I don't really think about it all that much actually. I stay out of areas I shouldn't be, I avoid situations that I shouldn't be in

And other than that, I figure if something happens, it happens and it's my time. C'est la vie

1:33 AM

Avoiding dangerous situation is your number one self defense technique. Most reliable thing going. Item two is a willingness to employ sneaker-do and get the heck out of dodge.

Yup, there's no shame in avoiding confrontation. And always hanging out with a friend who is slower than you are :)

I don't have to out run the bear...

Most of the martial artist I know--even the guys who are slick as ice on the mat--are very big on avoiding situations. Too many variable, too much risk, no reward.

Q: Delay in awarding two badges for same event

4:07 AM

Came across a little piece of fiction published in Nature: Tony Ballantyne's "If Only ...". It's short and I suspect that most Physics SE users will appreciate it.

Physics Meta

8:00 AM

0

Just today, I answered an interesting question about the intensity of sunrise in Australia vs Asia. When my accepted answer hit the 10 upvotes, I received ("instantly", or so it seemed) the "nice answer" badge. But since it was an accepted answer, I would have thought I hit the conditions for the...

discussion

10:07 AM

Hi,

Does anyone know the story behind the user "Ron Maimon"? I've seen a couple of his answers recently, but currently he seems to have left the site (suspension)

Some of his answers are really good.

1:23 PM

@Phonon Ron Maimon left of his own volition. It was about the same time I joined I think. If I recall (and there's a good chance I don't), he left because he was fed up with the politics of the site and several of the inner workings. And yes, his answer are usually decent

@Jim ah okay, thanks for the info.

2:52 PM

@Phonon If you're really interested, search the Meta site for his name & you'll see some details (Mods & SE employees keep a lot of details private as a matter of policy). He (along with former users Dilaton & Dimension10) runs PhysicsOverflow, which is supposed to be like MathOverflow but for physics.

If you browse PhysicsOverflow, you'll note that they take Physics.SE questions & answers and repost them there.

It's all legal (SE is CC-by-SA), but it seems a liitle dishonest to me because it inflates their question content

ah I see

yeah I understand, very strange

but what's their purpose?

if they're just re-doing what's already been done in physics.se

They feel Physics.SE is rejecting "perfectly legitimate & advanced theoretical questions" on the basis that it "looks" like homework

Plus they also like to be rude

wow...

I'm not kidding about that last comment too

such nonesense

yeah I know man

3:02 PM

They honestly feel that informing someone is wrong should not be done in a civil manner, that they should insult the person

hahahahaha

yeah 0 tolerance

so they just bash people right off

and how is the user-base number at the moment?

do they have more users already than Physics.SE does?

@Phonon As far as I can tell, they reserve it for crackpots & not the general public

:))

then we may still be in the majority :D

@Phonon I sincerely doubt it, but their true numbers are entirely hidden because they import Physics.SE users into their user-base, even if that person doesn't actually contribute

yep exactly

well at least I can say I have some idea what's going on

thanks for the info

3:07 PM

No problem

good to know these things, specially since I'm new to physics.sE

3 weeks ago finally decided to make an account and starting contributing as well, before I was always being led here by google

(starting to contribute*), I have to run off now, cya later mate.

Bye

3:56 PM

speaking of the importing users from SE, overflow has a profile for me with some of my answers (no surprise) but also they have some questions attributed to it that I never asked

There must be a second Jim. I must find him and destroy him

@Jim There's more than just another Jim:

physics.stackexchange.com/…

omg theres 13

Yep

Bet you one of those cats asked a Q that was imported over

Yep. But I'm not worried. I have more rep than all of them combined. I think that makes me the king of the Jims

Haha

Is that a title worth having?

4:00 PM

Meh, it comes with a parking space and diplomatic immunity. Plus all the other Jims have to call me "My Liege" or something like that

If we all changed our names to Jim, would we become Jim.SE?

@Jim Well that very well may be worth it

No, I think Jim.SE answers questions about Jim

@ACuriousMind As long as we don't have to talk about Gym's, I'm okay with being Jim

I think we'd be Physics.Jim

Ok, no more talk about changing our names to all Jim. You're starting to make me scared for my title

I'm going to lunch. When I get back, there better not be 20 new Jims

4:04 PM

@Jim Challenge accepted. ;)

4:47 PM

Hmmm changing my name to Jim would help alleviate the Kyle problem a bit ;)

Q: Are antiparticles just particle-shaped holes?

0

If particles are simply regions of space where certain quantum fields have non-zero divergence, are anti-particles simply the corresponding regions of opposite divergence? This seems like the intuitive answer, especially when considering the process of annihilation. I have heard before that a...

antimatter

That's clever.

5:06 PM

@KyleKanos Isn't that what "positrons are holes in the Dirac sea" is about?

I'm an astrophysicist, I am not very much aware/educated in Many-body theory

Nothing to do with many-body, the Dirac sea was one of the bizarre interpretations the early quantum mechanics came up with when confronted with the strange results of QFT

Basically, yes, you can think of particles as holes in an otherwise filled space, but you can also just stick to particles as you naively imagine them

Yep, unawares of QFT too

The viewpoints are equivalent, and it's really one of these eternal "But what is the world REALLY like?" battles

I see

I do hydrodynamics, sometimes magneto- hydrodynamics.

And sometimes with a Fokker-Planck type equation thrown in there for S&Gs

Okay, not really for S&Gs, but for research work

user54412

5:18 PM

@Kyle (!Kanos) I've been very distracted by your asymmetric orbits question

@ChrisWhite I still got pinged :/

user54412

lol

user54412

I've written a short python program that animates the orbits, and one thing that occurs to me is that we're not exploring the full potential well anyway

Does anyone else see a guy excitedly raising his arms into the air when someone writes " lol "? I can't unsee it.

o/

^ that's waving arms

lol looks more look like

Which, I think, is an apt image

user54412

5:24 PM

It's kinda obvious in retrospect, but it took me a colored plot of the potential to see it. See for instance v0=1, q=0.7, Lz=0.05, R0=0.3, z0=0.1, vR0=1, vz0=0.1 for time 200:

user54412

3

user54412

even the orbits that eventually fill space symmetrically don't reach everywhere kinetically allowed, so it's not unthinkable that there could be separate islands in phase space

user54412

each one might be filled ergotically, but they are dynamically distinct

user54412

and then the symmetry is simply broken by the asymmetric initial conditions, which chooses an island

user54412

it's weird that the islands overlap, but that's just in the 2D projection onto (R,z)-space of the 4D phase space

6:06 PM

@ChrisWhite huh that's pretty interesting

you could do this with surfaces of section too (bit of an obscure/arcane technique, but it works here I'm pretty sure... let me dig up a plot)

those are the same parameter choices and same layout as the plots in my question

oh and those are $z=0$, $p_z>0$ surfaces of section (the points)

user54412

6:30 PM

Well, that's an asymmetrically placed island alright

user54412

I suppose one could take a plot like the top left one, overplot the results from the normal ICs and the z-reversed ones in different colors, and slowly vary the parameters toward the bottom left one

user54412

And probably the ring will at some point bifurcate and all of one color goes up while the other goes down

user54412

The bottom right plot reminds me of those KAM theory/invariant tori we once touched on in a dynamics course (wish I had studied that more)

6:49 PM

@Jim hahaha I just got to read your messages from earlier Jim :D

yeah I had no idea about this physicsoverflow site until Kyle told me about it today

@ChrisWhite I think I'm missing something about your plots... are there supposed to be 2 line colours?

user54412

@Kyle I'm just thinking if you redid the upper left or upper right surface of section, but flipped the sign of z_0 and \dot{z}_0 (or just reversed your p_z>0 criterion I guess), you'd get points along the same curve. But if you did it in the lower left plot, you'd get points along a curve that mirrored the one shown across v_R=0.

@Jim it is a pity mainly to see that people are stumbling and wasting time over such silly things instead of just doing physics...what is we are all here for anyway.

user54412

But surely one can smoothly vary the parameters and initial conditions from one plot into the other, so there must be some sort of bifurcation (defined loosely).

@ChrisWhite I agree

but was wondering about your plot... you're saying something about colours, but I only see one coloured line (and two backgrounds, but I understand what those are showing)

user54412

6:57 PM

Oh sorry - I didn't plot different lines myself

user54412

That's just the one orbit, with the Phi>E and Phi<E regions of space

ah ok

I wonder if I could get my code to spit out a bunch of surfaces of section - it's old, from back when I was only a couple weeks into learning python, so it's very... well, 'bad' is the word that comes to mind

user54412

I'm tinkering with mine now - it's not particularly pretty (or fast)

I'm fussing with some actual work right now, but might tinker a bit later today

9:26 PM

Anybody around? I have some questions about interpreting something

lurking, what's up?

I'm using VTK to compute the volume given a shell (generated by an isosurface). In the code, it computes volume and then the "volume projected on each axis" and I'm unsure what that projection represents.

In the code it does:

xavg = (x[0] + x[1] + x[2]) / 3.0;

vol[0] += (area * u[0] * xavg);

vtk?

Where u[0] is the x componet of the normal vector on the shell. So it's adding up the area times the normal times the average X value of the triangle on the surface

nevermind, googled

9:29 PM

And calling that the "projection of the volume"

But I'm not sure what exactly that represented in physical terms.

ok so still dissecting the code... what is x[i]?

x coordinate

of...?

The surface is represented by triangles, so it's looping over each triangle

On the surface

you have a shell, it doesn't have a unique coordinate?

ah hm ok

9:33 PM

Finding the X coordinate of the centroid

Is what xavg is

so (xavg, yavg, zavg) is the centroid of the triangle?

Right

And u[0] = n_x, u[1] = n_y, u[2] = n_z, the components of the normal vector of the surface

And area is the area of the triangle

is u a unit vector?

Yes

I just can't picture what this is a measure of

hm

area * u[i] is the vector area

9:36 PM

Right, and xavg is the coordinate of the centroid, but I don't know what that matters

yeah so I'm now caught up with what is actually represented, trying to think of a geometrical meaning

are the volume and these 'volume projections' related somehow? maybe v^2 = vx^2 + vy^2 + vz^2?

Yes -- the comments say:

// Volume under triangle is projected area of the triangle times// the average of the three z values

It's using the Maximum Normal Unit Component method of computing the volume, so the total volume is:

this->Volume = (kxyz[0] * vol[0] + kxyz[1] * vol[1] + kxyz[2] * vol[2]);

Where the kxyz are called " weighting factors for the maximum unit normal component (MUNC)"

ok I think I get it

now to try and use words :/

if I take an arbitrary triangle and draw 3 lines in the xhat direction joining the vertices and the y-z plane, I outline a solid, a sort of triangular prism with crooked ends

I think that's the volume being computed for each triangle, and you can repeat the exercise along the other 2 axes for the other volumes

Okay, so when you sum that up over the entire surface, what "volume" is that? Or is there no real interpretation to it, other than it's added together with a weighting factor?

I think all three are valid approximations of the volume?

but you get greater accuracy by combining them in a particular way with the weights?

9:50 PM

That could be

note that this is pretty speculative

Much better than what I came up with

if you look at it, vol[0] has dimensions of volume

(weights are dim-less right?)

Yeah, they are

yeah thinking about it this is right

if your mesh has some particular symmetries then v = v[0] = v[1] = v[2], but in the more general case v[0], v[1] and v[2] will either over or under-estimate depending on the details

9:52 PM

Yeah, in some cases v_z is negative

But the others are positive and total volume is positive

I'm guessing for a spherical surface, all three would be equal

what do you mean by spherical surface?

you have a mesh of points, at best you have an approximation to a sphere

If your triangles made a sphere

Right. In the limit you actually got a sphere

So small triangles

yeah

better test would be a cube

or maybe not... the cube is kind of trivial

reducing to 2 dimensions, the picture I sort of have in mind is like this:

dot being the centroid, the red and blue regions are areas that carry some sort of information about the area of the triangle, combining them cleverly you can get an even better estimate

all a bit heuristic of course

Yeah, makes sense. I was kinda hoping it would be a measure of the size in X somehow

Cause the surface is really an isosurface in (x,y,t) space and I need some measure of it's size in (x,y) space. Like a maximum area, or average area, or... something along those lines of a cross-sectional plane

10:18 PM

does the volume projected to an area on x-y not work for you?

like, if you took all your (x,y,t) points and plotted y vs x, you could just make some measurement of that set of points?

Perhaps. I'll have to play with it. Unfortunately finding the isosurfaces and stuff is very expensive

So "playing" with the data is not cheap. And I haven't been able to produce a similar, smaller dataset yet

can't you just compute an isosurface and store it?

I tried to do that, but pickling the VTK object didn't work and I didn't realize it until it ran for awhile :/

So I'm trying to salvage what I have access to before I have to go back and re-run the analysis for 3-4 days